<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Lora on Yonk-Labs</title><link>https://yonk.dev/tags/lora/</link><description>Recent content in Lora on Yonk-Labs</description><generator>Hugo -- gohugo.io</generator><language>en</language><copyright>© 2026 Yonk-Labs</copyright><lastBuildDate>Mon, 13 Apr 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://yonk.dev/tags/lora/index.xml" rel="self" type="application/rss+xml"/><item><title>Training Sets for LoRA: How to Teach a 4B Model to Write Postgres SQL Without Crying</title><link>https://yonk.dev/blog/training-sets-for-lora-nl2sql/</link><pubDate>Mon, 13 Apr 2026 00:00:00 +0000</pubDate><guid>https://yonk.dev/blog/training-sets-for-lora-nl2sql/</guid><description>A layered training corpus — domain pairs, public ballast, and a reusable Postgres syntax corpus — is 80% of the work for a NL2SQL LoRA. The training config is YAML and patience.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://yonk.dev/blog/training-sets-for-lora-nl2sql/feature.jpg"/></item></channel></rss>